School-testing study provides valuable lesson

School test results, as we all know, depend on the neighbourhood. Students with affluent, educated parents will do better as a rule than those with poorer, less educated parents. North Toronto Collegiate will have better results than Parkdale. So, as a way of telling how good a school is at teaching, standardized tests are essentially useless. That, at any rate, is what teachers unions and other test haters claim.

An important new paper from the C.D. Howe Institute shows how simplistic and misguided that thinking is. Economics professor David Johnson of Wilfrid Laurier University has been studying results from EQAO, provincewide tests in reading, writing and mathematics that all students take in Grades 3, 6, 9 and 10 under the Education Quality and Accountability Office. He uses census data and other information about students' socioeconomic background to predict how certain schools would be expected to do on Grade 3 and 6 tests. Then he compares those predictions against how the schools actually performed.

Not surprisingly, schools with students from more affluent backgrounds tend to do better, but within that generality Prof. Johnson says there are sharp variations. Based on the above-average affluence of their students, Mount Hope Public School in Hamilton and St. Cecilia Catholic School in Toronto were each predicted to have a pass rate of 5 per cent above the provincial average. In fact, Mount Hope's pass rate in Grade 3 testing wound up 10.7 per cent below the provincial average while St. Cecilia was 22.3 per cent above. Obviously, St. Cecilia is doing something right. Mount Hope has some thinking to do.

Prof. Johnson found the same variations in schools with less privileged students. Based on its socioeconomic makeup, Cornell Junior Public School in Toronto should have had a pass rate 7.9 per cent below the provincial average. Instead it scored 15.1 per cent above. It is way, way ahead of other schools with students of similar background. What is Cornell doing to give its underprivileged students a leg up?

It's a legitimate question, but the teachers unions would rather keep us in the dark. Though they represent a profession dedicated to the pursuit of knowledge, they seem to feel that information of the kind being unearthed by Prof. Johnson is a dangerous thing.

In fact, it is priceless. It shows, to begin with, that background is not destiny. Kids in poor districts don't have to have lousy schools. Kids in rich areas don't always get great schools either.

"Anyone who claims schools with same background get the same results is wrong," says Prof. Johnson. "The data just screams this at you. The upside is that the quality of instruction offered by teachers and principals and education assistants matters - it makes a difference."

No matter what neighbourhood they are in, schools can turn themselves around. Armed with the EQAO results, many are doing just that. Principals pore over the school-by-school results when they come out every year. Why did my school score so much worse in Grade 6 math than the one three blocks over? Do we need to bring in a tutoring program or stage a math month?

When EQAO results in 2000 showed that just 49 per cent of Grade 3 students were meeting the provincial reading standard, schools ramped up their early reading programs. As a result, 61 per cent of students met the standard in 2008.

Parents scrutinize the scores, too. Does my daughter need extra help with writing? Why is my kid's school posting such poor results? Would the other, more successful neighbourhood school be better for her?

Test haters say this puts schools in competition with each other. Well, precisely. Students compete against each other all the time - in sports, for academic prizes, for class representative. Why should their schools be any different? Schools that do well in testing get a boost for their efforts; those that do poorly get a prod to do better.

By debunking the idea that school performance depends mainly on student background, Prof. Johnson has fired the starting gun on a healthy race to make all our schools as good as they can be - regardless of how rich or poor their students are.

HIGHS AND LOWS

The 10 best and worst Toronto public and Catholic elementary schools, according to the report, with their percentile ratings for Grade 3 and Grade 6 test resultsbased on averages from the past three school years. Top 10 schools have been placed in order of the highest cumulative percentile scores; bottom 10, in order of their lowest.